Re: [Libmesh-devel] Pthread Periodic Boundary Issue

[Derek G. <fri...@gm...>]

Nope - it's not that the PointLocator is NULL... the PointLocator is
_returning_ NULL (ie it couldn't find an element).

Look at line 1956 in fe_base.C

Derek


On Mon, Jul 8, 2013 at 9:28 PM, Roy Stogner <roy...@ic...>wrote:

>
> On Mon, 8 Jul 2013, Derek Gaston wrote:
>
>  It appears that something is wrong in there (or we're missing a lock
>> somewhere).  Here is what happens in optimized mode:
>> PeriodicBoundaries point locator object returned NULL!
>> Stack frames: 5
>> 0: 0   libmesh_oprof.0.dylib               0x0000000102d98390
>> libMesh::print_trace(std::**ostream&) + 64
>> 1: 1   libmesh_oprof.0.dylib               0x0000000102e3b6a4
>> libMesh::FEGenericBase<double>**::compute_periodic_**
>> constraints(libMesh::**DofConstraints&, libMesh::DofMap&,
>> libMesh::PeriodicBoundaries const&, libMesh::MeshBase const&,
>> libMesh::PointLocatorBase const*, unsigned int,
>> libMesh::Elem const*) + 10308
>>
>
> Wait - so NULL is getting passed in as that PointLocatorBase*?  That
> should *never* happen.
>
> If you could trace that back any further (does the sub_point_locator()
> call in ComputeConstraints fail? the sub_point_locator_call() in
> create_dof_constraints?) then we're getting much closer to the root of
> the problem.
> ---
> Roy